Make tensor tests to pass ‘float32’ dtype argument explicitly

[happyhj]

The goal of this PR is to fix duplicated test logic on tensor test.

• Adds explicit float32 dtype parameter to tensor tests.
  • There were tests for default dtype and float32 dtype, but they don't passing a dtype argument so both default dtype is used.
  • If the test is for the non-default parameter (like float32), then dtype should be passed explicitly.
• Adds dtype variation test to existing tensor tests which lack of tests for dtype variation.

---

For repository owners only:

Please remember to apply all applicable tags to your pull request.
Tags: FEATURE, BREAKING, BUG, PERF, DEV, DOC, SECURITY

For more info see: /~https://github.com/tensorflow/tfjs/blob/master/DEVELOPMENT.md

---

This change is 

[nsthorat]

Reviewed 1 of 1 files at r1.
Reviewable status: 0 of 1 approvals obtained

[nsthorat]

Thanks for the cleanup and addition of tests!